Affiliations: [a] School of Information Science and Engineering, Yunnan University, Kunming, Yunnan, China
| [b] School of Information Science and Technology, Yunnan Normal University, Kunming, Yunnan, China
Correspondence:
[*]
Corresponding author. Dan Xu, School of Information Science and Engineering, Yunnan University, Kunming, Yunnan, 650500, China. E-mail: danxu@ynu.edu.cn.
Abstract: Multiresolution technique is one of the most efficient approaches to improve the rendering performance, but its design and implementation for massive meshes are still very difficult. This paper gives a solution to the construction and rendering of multiresolution representation for massive meshes. Our approach provides a dual hierarchy: a cluster hierarchy for coarse-grained selective refinement and a vertex hierarchy for progressive meshes contained in the cluster for fine-grained local refinement. In order to promote the speed of the construction and rendering of multiresolution representation for massive meshes, we implement parallel half edge collapse to build the vertex hierarchy in the clusters and view-dependent refinement for vertex hierarchy in parallel with GPU. To prevent data transmission delay to stall the rendering, a separate CPU thread is used to prefetch clusters data into main memory. Our construction process is faster than a corresponding sequential implementation and the rendering results are of comparable quality.
